The X-Ray View of AGN Ionized Outflows

Yair Krongold (IA-UNAM)Elena Jimenez-Bailon (IA-UNAM)

Martin Elvis (CfA) Nancy Brickhouse (CfA) Luc Binette (IA-UNAM) Smita Mathur (OSU)

The X-Ray View of AGN Ionized Outflows

Fabrizio Nicastro (INAF/FORTH/CfA)

6/18/2008 AGN Workshop 2008 (F. Nicastro; Aghios Nikolaos, Crete) 2

Warm Absorbers = AGN Winds

•Winds with velocity = 500-3000 km s-1 ~1lt-day/yr

•Found in 50% of Seyfert 1s, PG Quasars (Reynolds,

1997; George+98, Piconcelli+05)

•Related to UV ‘narrow absorption lines’ - NALs•Possibly Ubiquitous in AGNs


Simple solution: 2-3 components

1 keV

NGC 3783 Chandra MEG 900 ksec exposure (Krongold+03)


• Similar kinematical properties

• No drag forces

• NGC 3783, NGC 985, IRAS 13349, NGC4051

• Also NGC 5548 when each Vel. Comp. is modeled independently

AGN X-Ray Outflows:2-3 phases in pressure equilibrium?


WA Open Questions

• Where do they arise? 106 in r! - Disk wind/torus/NELR• What is the geometry? Spherical/Bi-Conical/funnel• Relation with other AGN components? BELR, BALs, Torus• What are their Physical Properties?

Mass and KE outflow rates depends critically on wind location (0.01-1000 Maccr)

• Cosmologically important? Feedback, Co-evolution

Time Evolving Photoionization provides insights


Problem Arises from:Observables

Un-Observables

cnRQ

4π=

nn

Ugas

ph= 2


Solution: Time Evolving Photoionization

Response is not instantaneous:‘Ionization time’ and ‘Recombination time’

Step function change

Gradual WA response

Constrains ne independently on U, and so constrains R


XMM-Newton Observation

•103 ksec• >10x flux changes• EPIC + RGS•High S/N + R

Modelled with PHASE (Krongold+03)

Bound-free edges only

Full PHASE model

Following the Opacity of the WA in NGC 4051


NGC 4051 RGS Data

• Simple solution

• only 2 absorbing components (LIP and HIP)

HIPHIP

FF66--3535=4.4x10=4.4x10--1111 cgscgs ~ XMM~ XMM

LIPLIP


Similar Solutions RGS and EPIC

Allow us to constrain ionization Parameter in

EPIC data


NGC 4051:

2 WA components in photoionization

equilibrium(Krongold et al. 2007)

log Ux(t), measured

log Ux(t), predicted

from photoionization

equilibrium

Both WA components are DENSE and COMPACT

XMM EPIC Light Curve


EPIC Data

4X flux increaseRGS Data

Comparing HS and LS data

First Noted by Ogle et al. 2004


Robust Estimate of (neR2)

)4log()]4()/)(log[(

)4log()(log)(log22

2

RcnDStCtr

RcntQtU

eeff

eXX

ππ

π

−×=

=−=

Continuous Flow Model Ruled OutContinuous Flow Model Ruled OutThe Winds must be CompactThe Winds must be Compact

(neR2)LIP = 6.6 1039 cm-1 (neR2)HIP = 3.8 1037 cm-1


Lower limit on LIP ne and R

• Low Ionization Phase (LIP):in equilibrium at all times

teq(LIP) < Δtl,m < 3 ks

ne(LIP) > 8.1 107 cm-3

But (neR2)LIP = 6.6 1039 cm-1

R(LIP) < 8.9 x 1015cm < 0.0029 pc

R(LIP) < 3.5 light days


HIP ne and R

• High Ionization Phase (HIP):

out of equilibrium at extreme fluxes (high and low)

teqi,j+k(HIP) > Δtj+k > 10 ks

(recombining)

HIP in eq. at moderate fluxesteq

l,m(HIP) < Δtl,m < 3 ks

R(HIP)R(HIP) = (1.3= (1.3-- 2.6)x 102.6)x 101515 cm cm = (0.5= (0.5--1.0) light days1.0) light days

ne(HIP)=(0.6-2.1)x 107 cm-3

logU

x(t)


Warm Absorber (wind) Location

• Rules out Obscuring molecular torus – Minimum dust radius, rsubl(NGC4051) ~ 12-170 light-days

• RLIP < 3.5 ld; RHIP~0.5-1 ld: consistentSpatially co-located ?

Disk winds, on BELR scale

lightlight--daysdays 5510101515

~~HeIIHeIIBELRBELR

HHββBELRBELR

Dusty molecular torus

LIP

Dust sublimation radius

HIP

~ 52000 Rg ~ 3000 Rg


Cylindrical/Conical Geometry

• All Spherical configuration related to known structures are ruled out. • Thin spherical shells are still possible, but implausible (need fine-tuning not

to degenerate into a continuous flow: which is ruled out): testable by re-observing NGC 4051

• Next simpler symmetry: cylindrical (or bi-conical): consistent with all our findings

Ṁout = (0.02-0.05) Ṁaccr

ΔR(δ)

ϕ

ΔrR(δ)

δ

dr=vdt


Mass Outflow Rates vs Mass Accretion Rate for NGC 4051

ϕδ

δϕπ

sincos)]/(2)/[(

)]cos(/)[(8.022

2

RRRR

vRnmM repout

Δ+Δ×

×−=&

For ϕ=900 and δ=300 we measure: ṀLIP < 0.9 x 10-4 M๏ yr-1 = 0.02 ṀaccrṀHIP = (0.7-1.4) x 10-4 M๏ yr-1 = (0.02-0.03) Ṁaccr

Ṁout = (0.02-0.05) Ṁaccr

KP = (3-8)x1037 erg s-1

(assuming v = 2vr = 1000 km s-1)


AGN-Feedback from NGC 4051

Assuming MBHNGC4051~ 2x106 M◉ all accreted

Mout = (0.02-0.05)MBH = (0.4-1) x 105 M◉Unimportant for IGM metal-feeding

KEaval ≈ (0.4-1) x 1053 ergs≲ 1055 ergs required to unbound hot ISM and inhibit large-scale star formation (e.g. Hopkins+06)

May control star-formation<< ~1060 ergs required to control host-galaxy and surrounding-IGM evolution (e.g. Natarajan+06)

Unimportant for MBH-σ

But NGC 4051 LOW MBH and LBOL


• 2 Month Monitoring •227 ksec•1 Visit every week

•High S/N

•x4 flux variation

SUZAKU Monitoring Campaign of NGC 5548

6/18/2008 AGN Workshop 2008 (F. Nicastro; Aghios Nikolaos, Crete) 21800 ksec MEG & LEG exposure

(Andrade-Velazquez+08: SEE POSTER)

WA properties constrained through Chandra HR data

2 Velocity Components

Each modeled with 2 absorbers

HV: V~1100 km/sSHIP & HIP

LV: V~500 km/s HIP & LIP


Suzaku Light Curve

SHIP (HV)

HIP

(HV+LV)

LIP (LV)

log Ux(t)

measured

log Ux(t)

predicted at PE

SHIP does not

respond

HIP in PE only

in average flux

LIP in PE

Always

NGC 5548 Absorbing Components


Wind Locations and Mass Rates(Assuming NGC 4051 empty bi-conical geometry)

SHIP (HV)teq > 2 Msec

ne < 5.4 x 105 cm-3

R > 0.2 pc~ 30000 RG

HIP(HV+LV)teq ~ 0.5-1 Msec

ne < 0.8-2.6 105 cm-3

R ~ 0.7 – 1.6 pc

LIP (LV)teq < 0.6 Msec

ne > 1.0 x 105 cm-3

R(LIP) < 3.1 pc

Much more massive & energetic than NGC 4051

Ṁout > 2.5 Ṁaccr (0.2 M๏ yr-1)

KP > 1.2 x 1041 ergs s-1 (assuming v = 2vr = 1000 km s-1)


AGN-Feedback from NGC 5548

Assuming MBHNGC5548= 7x107 M◉, all accreted

Mout = ~ 2.5 MBH = 1.8 x 108 M◉Metal-feeds the IGM

KEaval ≈ 3.3 x 1057 ergs>> 1055 ergs required to unbound hot ISM and inhibit large-scale star formation (e.g. Hopkins+06)

Unbinds hot ISM and stop star-formation<< ~1060 ergs required to control host-galaxy and surrounding-IGM evolution (e.g. Natarajan+06)

Unimportant for MBH-σ

NGC 5548 still LOW MBH and LBOL


Mout ≥ 2.5 MBH ≈ 109-1010 M◉ !!! (neglecting ISM mass entraining)

Comparable with ULRIGS (+ QSO Envs are Metal-rich)

Important for IGM metal-feeding

KEaval ≈ few x 1059 ergs !!! (neglecting wind acceleration)

Comparable to Ebound(Bulge) and E from simulations

Controls star-formation + Regulates MBH-σ

Speculative !!! More data needed!

Extrapolating to Massive QSOs(MBH ≈ few x 109 M◉, Lbol≈1047 ergs s-1

=> Ṁaccr ≈ 15 M๏ yr-1, for η=0.1)


Conclusions• AGN X-ray Outflows are:

– Disk winds (NGC 4051 + NGC 5548)– Dense, compact, and multi-phase (possibly in pressure balance)

• Location & Geometry:– NELR and Torus are ruled out as Originating locations (too far)– Continuous Flow is ruled out (WA are seen varying) – Spherical geometry is (almost) ruled out – Conical (cylindrical) geometry works:

• NH down cone > 10 x NH observed, possibility NALs=BALs (Krongold et al., 2007)• Consistent with transverse motion evidence (UV data: REF.)

• Feedback: NGC 4051 + NGC 5584 results imply (speculative!!!): – Large Mass and Energy deposited into Bulge/IGM, if scales with Lbol– Important for IGM Metal-feeding– Control Start-Formation– Regulate MBH-σ

• More Objects needed

Documents

The X-Ray View of AGN Ionized Outflows